|Publication number||US4736254 A|
|Application number||US 06/800,439|
|Publication date||Apr 5, 1988|
|Filing date||Nov 21, 1985|
|Priority date||Nov 22, 1984|
|Publication number||06800439, 800439, US 4736254 A, US 4736254A, US-A-4736254, US4736254 A, US4736254A|
|Inventors||Hiroaki Kotera, Motohiko Naka|
|Original Assignee||Matsushita Electric Industrial Co., Ltd.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (10), Referenced by (58), Classifications (6), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention relates to the generation of a halftone signal by comparison between an original gray scale value and a threshold value having a prescribed ordered-dither distribution.
Facsimile receivers and dot printers are of the type wherein the printing of dots is essentially an all or nothing binary process. While this all or nothing process is desirable for printing alphanumeric characters or other symbols, it is disadvantageous for printing images having gray scales or light gradations in continuous tones. This problem has been solved by transforming the continuous tone of the original image into halftone or pseudo halftone images using a mesh distribution or ordered-dither distribution known as "dither". Pseudo halftone images comprise a varying number of dots corresponding to the shades or tones to be reproduced. When the dots are small compared to the visual sharpness of the human eye, they are subliminal to the eye and are not recognized. Thus, dots and spaces visually fuse together on a writing surface to present a continuous tone.
One typical example of pseudo halftone generators as shown and described in U.S. Pat. No. 4,032,978 comprises a pseudo-random number generator and a comparator that compares input gray scale levels Xij (where i and j are representative of the row and column positions of a matrix array, respectively, in which dots and spaces are indicated) with the output of the pseudo-random number generator to produce a print signal Gij depending on whether the gray scales are higher or lower than a "dithered" reference Dij represented by the pseudo-random number. The reference Dij is varied according to a prescribed pseudo-randomness, or probability of occurrences in the directions of the rows and columns so that the output Gij has a distribution in which the ratio of dots and spaces varies with the prescribed randomness.
It is desirable to provide different distributions of dots and spaces according to different shades or optical density to improve continuous tones. However, the prior art halftone generator is not satisfactory in terms of the degree of freedom for selecting a desired dither distribution of threshold values.
It is therefore an object of the present invention to provide an apparatus and method that permits the determination of a halftone value with a freedom of choice from among different ordered-dither threshold distributions according to different optical densities, or gray scale values.
This object is obtained by dividing a matrix array of a memory into a plurality of identical "imaginary" matrices and storing threshold values in particular cells of each imaginary matrix according to a prescribed ordered-dither distribution.
According to the present invention, there is provided an apparatus for generating a print signal having one of two discrete levels for each of print positions along each of a plurality of print lines depending on a gray scale value of an original relative to a threshold value. The apparatus comprises a memory having an αM×βN matrix array of cells each storing a particular threshold value, where M, N, α and β are integers, at least one of α and β being greater than unity and at least one of M and N being greater than unity. The matrix array is divided into d×B identical imaginary matrices of M×N elemental areas each containing a cell, in particular one of α×β positions. The threshold values of each of the imaginary matrices have a prescribed ordered-dither distribution on the plane of each imaginary matrix. A first counter is incremented in response to each print position for addressing the memory in a given direction and a second counter is incremented in response to each print line for addressing the memory in a direction normal to the given direction to sequentially read the stored threshold values from the memory for each print position. A comparator is provided for comparing the gray scale value of the original with each of the threshold values read out of the memory and generating a print signal representative of one of the discrete levels.
Preferably, the threshold values which are stored in the cells of positionally corresponding elemental areas of the imaginary matrices form a group of successively varying threshold values for each elemental area according to a prescribed numerical relationship between the imaginary matrices.
The present invention will be described in further detail with reference to the accompanying drawings, in which:
FIG. 1 is a block diagram of a pseudo halftone generator embodying the present invention;
FIG. 2 is a digrammatic representation of an ordered-dither distribution of threshold levels useful for describing the underlying principle of the invention;
FIG. 3 is a diagrammatic representation of a dot-dispersive ordered-dither distribution of threshold levels based on a combination of a dot-dispersive Bayer distribution on the two-dimensional planes of plural imaginary matrices and a dot-concentrative distribution between different imaginary matrices;
FIG. 4 is a diagrammatic representation of a dot-concentrative ordered-dither distribution of threshold levels based on a combination of a mesh distribution on the two-dimensional planes of plural imaginary matrices and a dot-concentrative distribution between different imaginary matrices;
FIG. 5 is a diagrammatic representation of a dot-dispersive ordered-dither distribution of threshold levels based on a combination of a mesh distribution on the two-dimensional planes of plural imaginary matrices and dot-dispersive distribution between different imaginary matrices;
FIG. 6 is a diagrammatic representation of a dot-concentrative ordered-dither distribution of threshold levels based on a combination of a dot-concentrative spiral distribution on the two-dimensional planes of plural imaginary matrices and a dot-dispersive distribution between different imaginary matrices;
FIG. 7 is a graphical representation of optical density as a function of threshold levels;
FIG. 8 is a diagrammatic representation of threshold levels based on a combination of Bayer and mesh distributions on the two-dimensional planes of different imaginary matrices and dot-dispersive and dot-concentrative distributions between different imaginary matrices; and
FIG. 9 is a diagrammatic representation of threshold levels based on a combination of spiral and Bayer distributions on the two-dimensional planes of different imaginary matrices and a dot-dispersive distribution between different imaginary matrices.
Pseudo halftone as discussed herein refers to the process of reproducing continuous tone images or gray scales with a printing device having only two levels of gray, namely, black and white, and operating with a fixed printed dot size and fixed dot-to-dot spacing. An embodiment of the pseudo halftone generator of the present invention is illustrated in FIG. 1. In the embodiment shown, a quantized, binary coded gray scale input is supplied to a line buffer memory 1 on lines 2. The gray scale input typically comprises 6-bit words each representing the gray scale level of a pixel or "dot" image of the original halftone picture, so that it can be represented with a resolution of sixty-four (=26) levels of gradation. Digital bits of a full print line are stored into the memory 1. A clock pulse is applied to a horizontal address 11, or "row" counter 3 on line 4 in synchronism with the application of each 6-bit word to the buffer memory 1. Address codes are successively generated by counter 3 and supplied on lines 5 to buffer memory 1 for addressing successive storage locations of the memory to read the stored 6-bit words out of it into a digital comparator 6. Comparator 6 provides a comparison between each 6-bit word from memory 1 and a 6-bit word read out of a threshold table memory 7 which represents an ordered-dither threshold variation. If the digital value of the output of line memory 1 is lower than the threshold, comparator 6 supplies a logical 0 to a dot printer 10 and if it is higher than the threshold comparator 1 supplies a to the dot printer.
For purposes of illustration, table memory 7 comprises cells arranged in a matrix array of 8 rows and 8 columns. Horizontal address counter 3 increments a three-bit "row" counter 8 upon the completion of a print line to cause it to generate a "row" address for the table memory 7. The column address of the memory 7 is generated by a three-bit "column" counter 9 by counting the clock pulses on line 4. Six-bit threshold representative words are serially read out of memory 7 in the direction of its rows in synchronism with the six-bit data words from buffer memory 1 and supplied to digital comparator 6. Threshold data stored in each row of the memory 7 are therefore repeatedly read out of it into comparator 6 in response to row counter 8 being incremented as a line print is completed. All the stored data are therefore repeatedly read out of memory 7 until the print position reaches the end of a full page. A series of pulses is generated by comparator 6 and fed to dot printer 10 to produce a pseudo halftone image on successive print lines.
FIG. 2 is an illustration of the basic architecture of the table memory 7. Memory 7 is divided into a plurality of identical "imaginary memory layers", or matrices 7-1, 7-2, 7-3 and 7-4. Each matrix is divided into sixty-four (=8×8) cells, as indicated by intersecting broken lines. Each cell corresponds to a dot on the print line and stores a particular 6-bit word representing a pseudo random threshold variable. Each imaginary matrix is further divided by intersecting solid lines into a 4-by-4 matrix of sixteen "elements". Each element stores in a particular one of its four positions a particular one of threshold values having an ordered-dither distribution on each imaginary matrix. The ordered dither variables Ckij (where k identifies the 4-by-4 matrices, or "phases", i and j represent the column and row positions of the 4-by-4 matrices) stored in the matrices 7-1 to 7-4 are given by:
Ckij =Clij +k-1 (1)
where, Clij represents the initial pseudo-random variable of phase "1" for each series of consecutive numbers. In the illustrated embodiment, C111 is assumed to be "1" and consecutive numbers "1", "2", "3" and "4" and upper-left, upper-right, lower-left and lower-right cells of the element at the first row and first column of the 4-by-4 matrices 7-1, 7-2, 7-3 and 7-4, respectively (or phases k=1, k=2, k=3 and k=4, respectively). With all the cells of the matrices being filled with variables Ckij, the "imaginary" matrices are overlaid one upon another to produce a composite matrix 7-5.
The ordered dither variables Ckij can also be given by:
Ckij =Clij +n(k-1) (2)
where, n is the number of elements in each of the 4-by-4 matrices.
The distribution of initial dither variables Clij on the first phase "imaginary" matrix is determined in accordance with an "ordered dither" pattern which is available in a number of conventional methods known as "Bayer", "mesh" or "spiral" distribution. It is found that with the ordered dither pattern according to Equation 1 there is a tendency for dots to concentrate. With the pattern according to Equation 2 there is a tendency for dots to disperse. Thus, Equations 1 and 2 can be used advantageously for different applications. For example, Equation 1 is used for applications where a high degree of gradation is desired. Whereas Equation 2 is used for applications where a high degree of image resolution is desired.
The present invention allows Equations 1 and 2 to be selectively combined with one of the Bayer, mesh and spiral distributions to generate particular halftone characteristics for different optical densities. The Bayer distriubtion is known to impart a dot-dispersive characteristic and the spiral distribution a dot-concentrative characteristic. The dot-dispersive and dot-concentrative characteristics are suitable for emphasizing image resolution and image gradation, respectively. The mesh distribution is known to impart a characteristic which is intermediate the characteristics of the Bayer and spiral distributions.
FIG. 3 is an illustration of a specific example of the invention in which Equation 1 (dot-concentrative) and the Bayer distribution (dot-dispersive) are combined to produce a dot-dispersive ordered-dither distribution having an improved image resolution characteristic. Threshold values in digital form (6-bit representation) are stored in those particular cells of each of imaginary matrices 73-1 to 73-4. The threshold values stored in each imaginary matrix have a Bayer distribution and those values stored in positionally corresponding elemental areas of the imaginary matrices have a distribution according to Equation 1 which is dot-concentrative.
FIG. 4 is an illustration of a second example in which Equation 1 is combined with a mesh distribution. Digital threshold values stored in each of imaginary matrices 74-1 to 74-4 have a mesh distribution and those values stored in positionally corresponding elmental areas have a dot-concentrative distribution according to Equation 1. The combination of Equation 1 and the mesh distribution results in a dot-concentrative characteristic which emphasizes image gradation.
FIG. 5 is an illustration of a third example in which Equation 2 (dot-dispersive) is combined in a manner identical to FIGS. 3 and 4 with a mesh distribution to provide a dot-dispersive characteristic which emphasizes image resolution. Threshold values stored in each of imaginary matrices 75-1 to 75-4 have a mesh distribution and those values stored in positionally corresponding elemental areas have a distribution according to Equation 2, which is dot-dispersive.
FIG 6 is an illustration of a fourth example in which Equation 2, which is dot-dispersive, is likewise combined with a spiral distribution (dot-concentrative) to provide a dot-concentrative characteristic which emphasizes image gradation. Threshold values stored in each of imaginary matrices 76-1 to 76-4 have a spiral distribution and those values stored in positionally corresponding elemental areas have a distribution according to Equation 2.
It is seen therefore that the matrix array of memory 7 can be defined as a set of k (=α×β) imaginary matrices each having N×M elemental areas of α×β cells each (where, N, M, α and β are integers, at least one of N and M being greater than unity, and at least one of α and β being greater than unity).
According to a practical aspect of the invention, a set of different ordered-dither imaginary matrices is prepared according to the contents of the original. In FIG. 7 in which the optical density of the original is plotted as a function of threshold values ranging from zero to 63, it is shown that image resolution is emphasized in highlight and shadow areas of the original and image gradation is emphasized on delicately varying shading areas having an optical density between low and high halftone values.
As illustrated in FIG. 8, imaginary matrices 78-1 and 78-4 (k=1, k=4) have a combination of a Bayer distribution on each of their planes and a distribution according to Equation 2 between positionally corresponding elemental areas of matrices 78-1 and 78-4 to give emphasized resolution to highlight and shadow portions of the original, respectively. Imaginary matrices 78-2 and 78-3 are formed by a combination of a mesh distribution on each of these imaginary matrices and a distribution according to Equation 1 between positionally corresponding elemental areas of matrices 78-2 and 78-3. Matrices 78-2 and 78-3 emphasize gradation on areas having low and high halftone values. Imaginary matrices 78-1 to 78-4 are combined to form an ordered-dither distribution matrix 78-5.
On the other hand if gradation is to be emphasized for highlight and shadow areas while resolution is to be emphasized on delicately varying halftone areas, it is appropriate to form a set of imaginary matrices as shown in FIG. 9. Imaginary matrices 79-1 and 79-4 are formed by a combination of a spiral distribution on each of these matrices and a distribution according to Equation 2 between positionally corresponding elemental areas of these matrices to impart a dot-concentrative characteristic to the highlight and shadow areas, respectively. Imaginary matrices 79-2 and 79-3 are formed by a combination of a Bayer distribution on each of these matrices and a distribution according to Equation 2 between positionally corresponding elemental areas of these matrices to impart a dot-dispersive characteristic to the halftone areas. Imaginary matrices 79-1 to 79-4 are combined together to produce a matrix 79-5.
A plurality of sets of different ordered-dither distributions are prepared for table memories and a desired one is selected for a particular purpose.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4032978 *||Jan 20, 1975||Jun 28, 1977||International Business Machines Corporation||Pseudo halftone print generator and method|
|US4517605 *||Feb 16, 1982||May 14, 1985||Canon Kabushiki Kaisha||Image signal processing apparatus|
|US4553173 *||Dec 20, 1982||Nov 12, 1985||Canon Kabushiki Kaisha||Image processing apparatus|
|EP0074794A2 *||Sep 8, 1982||Mar 23, 1983||Xerox Corporation||Apparatus and method for producing halftone patterns|
|JPS5799867A *||Title not available|
|JPS6059866A *||Title not available|
|JPS56117478A *||Title not available|
|JPS59189782A *||Title not available|
|JPS60113582A *||Title not available|
|JPS60116283A *||Title not available|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US4769713 *||Feb 26, 1987||Sep 6, 1988||Hosiden Electronics Co. Ltd.||Method and apparatus for multi-gradation display|
|US4823189 *||May 8, 1987||Apr 18, 1989||Tektronix, Inc.||Method and apparatus for creating a dither pattern|
|US4989096 *||May 16, 1989||Jan 29, 1991||Crosfield Electronics Limited||Producing half-tone images of higher resolution through interpolation of original data|
|US5041920 *||Nov 15, 1989||Aug 20, 1991||Xerox Corporation||Image halftone generation by static RAM look-up table|
|US5051841 *||Oct 16, 1989||Sep 24, 1991||Bowers Imaging Technologies, Inc.||Process for providing digital halftone images with random error diffusion|
|US5107346 *||Apr 13, 1990||Apr 21, 1992||Bowers Imaging Technologies, Inc.||Process for providing digital halftone images with random error diffusion|
|US5111310 *||Dec 4, 1990||May 5, 1992||Research Technologies Corporation, Inc.||Method and apparatus for halftone rendering of a gray scale image using a blue noise mask|
|US5150428 *||Apr 1, 1991||Sep 22, 1992||Eastman Kodak Company||Method for generating halftone image data with randomly selected threshold array|
|US5175807 *||Jun 7, 1991||Dec 29, 1992||Quantel Limited||Video signal processing with added probabilistic dither|
|US5196942 *||Sep 17, 1990||Mar 23, 1993||Xerox Corporation||System and method for determining picture element values|
|US5204753 *||Jun 5, 1992||Apr 20, 1993||Eastman Kodak Company||Multi-bit rendering method and arrangement for continuous tone picture representation and printing|
|US5258849 *||Jun 5, 1992||Nov 2, 1993||Eastman Kodak Company||Halftone dot arrangement in gray level halftone printing|
|US5258850 *||Jun 5, 1992||Nov 2, 1993||Eastman Kodak Company||Line screen design for gray scale rendering|
|US5291311 *||Jun 29, 1992||Mar 1, 1994||Eastman Kodak Company||Apparatus and method for generating multi-level output values for pixels in a halftone cell|
|US5297014 *||Jan 3, 1992||Mar 22, 1994||Canon Kabushiki Kaisha||Switching DC power supply apparatus|
|US5317418 *||May 11, 1993||May 31, 1994||Hewlett-Packard Company||Halftone images using special filters|
|US5341228 *||Dec 3, 1991||Aug 23, 1994||Research Corporation Technologies||Method and apparatus for halftone rendering of a gray scale image using a blue noise mask|
|US5365251 *||Aug 28, 1992||Nov 15, 1994||Xerox Corporation||Image quality improvement by hierarchical pattern matching with variable size templates|
|US5371515 *||Sep 28, 1989||Dec 6, 1994||Sun Microsystems, Inc.||Method and apparatus for non-linear dithering of digital images|
|US5404427 *||Dec 24, 1992||Apr 4, 1995||Quantel Limited||Video signal processing with added probabilistic dither|
|US5444551 *||Jul 22, 1992||Aug 22, 1995||Eastman Kodak Company||Process for specifying mean-preserving mutli-level halftone matrices with varying amounts of modulation|
|US5477305 *||Nov 22, 1994||Dec 19, 1995||Research Corporation Technologies||Method and apparatus for halftone rendering of a gray scale image using a blue noise mask|
|US5489991 *||Oct 9, 1992||Feb 6, 1996||International Business Machines Corporation||Method for generating digital halftone images|
|US5535020 *||Oct 15, 1992||Jul 9, 1996||Digital Equipment Corporation||Void and cluster apparatus and method for generating dither templates|
|US5602971 *||May 8, 1995||Feb 11, 1997||Agfa-Gevaert, Nv||Multilevel halftoning using a randomised bayer matrix|
|US5633729 *||Dec 22, 1994||May 27, 1997||Eastman Kodak Company||Processes for minimizing the quantization error in multi-level halftone transfer functions|
|US5689623 *||Mar 27, 1995||Nov 18, 1997||Optronics International Corporation||Spread spectrum digital screening|
|US5701366 *||Sep 4, 1996||Dec 23, 1997||Canon Information Systems, Inc.||Halftoning with gradient-based selection of dither matrices|
|US5708518 *||Apr 11, 1995||Jan 13, 1998||Research Corporation Technologies, Inc.||Method and apparatus for halftone rendering of a gray scale image using a blue noise mask|
|US5726772 *||Oct 6, 1995||Mar 10, 1998||Research Corporation Technologies||Method and apparatus for halftone rendering of a gray scale image using a blue noise mask|
|US5745259 *||Jan 29, 1996||Apr 28, 1998||Digital Equipment Corporation||Void and cluster apparatus and method for generating dither templates|
|US5784049 *||Dec 18, 1996||Jul 21, 1998||The Harlequin Group Ltd.||Image display using evenly distributed intensity clusters|
|US5801838 *||May 20, 1997||Sep 1, 1998||Array Printers Ab||Method and device to improve print quality of gray scales and color for printers|
|US5808622 *||Aug 6, 1996||Sep 15, 1998||Hall; Daniel||Image display using irregularly placed curving structures|
|US5815286 *||Feb 1, 1995||Sep 29, 1998||Dainippon Screen Mfg. Co., Ltd.||Method and apparatus for producing halftone image with threshold matrix|
|US5838462 *||Apr 1, 1996||Nov 17, 1998||Xerox Corporation||Hybrid imaging system|
|US5912745 *||Apr 8, 1997||Jun 15, 1999||Digital Equipment Corporation||Void and cluster apparatus and method for generating dither templates|
|US5923823 *||May 15, 1997||Jul 13, 1999||Dainippon Screen Mfg. Co., Ltd.||Method and apparatus for generating halftone dots|
|US6002493 *||Dec 13, 1993||Dec 14, 1999||Case; Robert M.||Method for reproducing an image|
|US6172773||May 10, 1999||Jan 9, 2001||Compaq Computer Corporation||Void and cluster apparatus and method for generating dither templates|
|US6262811 *||Jan 7, 1998||Jul 17, 2001||Xerox Corporation||Increased functionality for holladay halftoning|
|US6731407 *||Nov 1, 1999||May 4, 2004||Seiko Epson Corporation||Image processing method and device|
|US7701615 *||Sep 22, 2006||Apr 20, 2010||Brother Kogyo Kabushiki Kaisha||Image forming method and image forming apparatus|
|US8243335 *||Sep 30, 2008||Aug 14, 2012||Canon Kabushiki Kaisha||Device for changing screen ruling for image formation in accordance with relationship between luminance and saturation|
|US20040090654 *||Sep 15, 2003||May 13, 2004||Luc Minnebo||FM screening with sub-dot phase modulation|
|DE4244835A1 *||Dec 23, 1992||Jul 13, 1995||Josef Dirr||Telefax or colour television image coding system|
|EP0642258A2 *||Aug 16, 1994||Mar 8, 1995||AGFA-GEVAERT naamloze vennootschap||Frequency modulated halftone image and method for making same|
|EP0665673A2 *||Jan 31, 1995||Aug 2, 1995||Dainippon Screen Mfg. Co., Ltd.||Method and apparatus for producing a halftone image using a threshold matrix|
|EP0682438A1 *||Mar 31, 1995||Nov 15, 1995||AGFA-GEVAERT naamloze vennootschap||Multilevel halftoning using a randomised Bayer matrix|
|EP0710007A2 *||Oct 24, 1995||May 1, 1996||Xerox Corporation||Image halftoning system|
|EP0740457A1 *||Apr 28, 1995||Oct 30, 1996||AGFA-GEVAERT naamloze vennootschap||Improved halftone screen and method for making same|
|EP0910206A1 *||Apr 28, 1995||Apr 21, 1999||AGFA-GEVAERT naamloze vennootschap||Improved halftone screen and method for making same|
|EP1307041A2||Oct 8, 2002||May 2, 2003||Fuji Photo Film Co., Ltd.||Image halftoning using threshold data|
|EP1401189A1 *||Sep 17, 2002||Mar 24, 2004||Agfa-Gevaert||FM screening with sub-dot phase modulation|
|WO1991004864A1 *||Oct 4, 1990||Apr 18, 1991||Array Printers Ab||Method to improve printing quality of grey scales and colour for printers|
|WO1992010905A1 *||Dec 3, 1991||Jun 25, 1992||Res Corp Technologies Inc||Method and apparatus for halftone rendering of a gray scale image using a blue noise mask|
|WO1995017070A1 *||Dec 12, 1994||Jun 22, 1995||Robert Maxwell Case||Method for reproducing an image|
|WO1996004748A1||Jul 28, 1995||Feb 15, 1996||Harlequin Group Limited||Image display using evenly distributed intensity clusters|
|Cooperative Classification||H04N1/4055, H04N1/4051|
|European Classification||H04N1/405B, H04N1/405C|
|Feb 4, 1986||AS||Assignment|
Owner name: MATSUSHITA ELECTRIC INDUSTRIAL CO., LTD., 1006, OA
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNORS:KOTERA, HIROAKI;NAKA, MOTOHIKO;REEL/FRAME:004513/0878
Effective date: 19851120
|Sep 30, 1991||FPAY||Fee payment|
Year of fee payment: 4
|Sep 26, 1995||FPAY||Fee payment|
Year of fee payment: 8
|Sep 27, 1999||FPAY||Fee payment|
Year of fee payment: 12